Transcript ppt

Introduction to Threads
Race Conditions
1
Process Address Space Revisited
Stack
Stack
Stack
Heap
Heap
Data
Data
Code
Code
OS
OS
(a) Process with
Single Thread
(b) Process with
Two Threads
2
What are Threads?

Thread
- Independent stream of instructions
- Basic unit of CPU utilization

A thread contains
- A thread ID
- A register set (including the Program Counter PC)
- An execution stack

A thread shares with its sibling threads
- The code, data and heap section
- Other OS resources, such as open files and signals
3
Single and Multi-Threaded Processes
4
Multi-Threaded Processes (2)

Each thread has a private stack

But threads share the process address space!

There’s no memory protection!

Threads could potentially write into each other’s stack
5
Why use Threads?

A specific example, a Web server:
do
{
get web page request from client
check if page exists and client has permissions
transmit web page back to client
} while(1);

If transmission takes very long time, server is unable to
answer other client’s requests. Solution:
do
{
get web page request from client
check if page exists and client has permissions
create a thread to transmit web page back to client
} while(1);
6
pthreads (5.4)

Refers to the POSIX standard (IEEE 1003.1c)

API for thread creation and synchronization

Common in UNIX operating systems
7
Java Threads (5.8)

Java threads may be created by:
- Extending Thread class
- Implementing the Runnable interface

JVM manages Java threads
- Creation
- Execution
- Etc.
8
Race Conditions
9
Concurrent Threads




Concurrent threads come into conflict with each other
when they come to use shared resources
Atomic actions are indivisible. In hardware, loads
and stores are indivisible.
On a processor, a thread switch can occur between
any two atomic actions; thus the atomic actions of
concurrent threads may be interleaved in any
possible order.
Result of concurrent execution should not depend on
the order in which atomic instructions are interleaved.
10
badcnt.c: An Incorrect Program
#define NITERS 100000000
unsigned int cnt = 0; /* shared */
int main() {
pthread_t tid1, tid2;
pthread_create(&tid1,
count,
pthread_create(&tid2,
count,
NULL,
NULL);
NULL,
NULL);
pthread_join(tid1, NULL);
pthread_join(tid2, NULL);
if (cnt != (unsigned)NITERS*2)
printf("BOOM! cnt=%d\n",
cnt);
else
printf("OK cnt=%d\n",
cnt);
}
/* thread routine */
void * count(void *arg) {
int i;
for (i=0; i<NITERS; i++)
cnt++;
return NULL;
}
linux> ./badcnt
BOOM! cnt=198841183
linux> ./badcnt
BOOM! cnt=198261801
linux> ./badcnt
BOOM! cnt=198269672
cnt should be
equal to 200,000,000.
What went wrong?!
11
Critical Sections

Critical sections are blocks of code that access
shared data.
/* thread routine */
void * count(void *arg) {
int i;
for (i=0; i<NITERS; i++)
cnt++;
return NULL;
}

The objective is to make critical sections behave as if
they are atomic operations: if one process uses a
shared piece of data, other processes can’t access it.
12
Atomic Operations

The statement
cnt++;
machine
level
R = cnt
R = R + 1
cnt = R
must be performed atomically (cnt is shared)

An atomic operation is an operation that completes in its
entirety, without interruption
13
Race Conditions

One possible interleaving of statements is:
R1 = cnt
R1 = R1 + 1
<timer interrupt ! >
R2 = cnt
R2 = R2 + 1
in = R2
<timer interrupt !>
cnt = R1

Race condition:
The situation where several
processes access shared data
concurrently. The final value of
the shared data may vary from
one execution to the next.
Then cnt ends up incremented once only! Threads
overwrite each other’s data.
14
Practice Exercise

Consider two threads sharing a global variable count,
initially 10:
Thread A
count++;

Thread B
count--;
What are the possible values for count after both threads
finish executing:
15
Practice Exercise


Consider the following three concurrent threads that
share a global variable g, initially 10:
Thread A
Thread B
Thread C
g = g * 2;
g = g + 1;
g = g - 2;
What are the possible values for g, after all three threads
finish executing?
16
Hands-on Session

Complete the hands-on POSIX programming exercises
posted on the class website.
- Creating threads
- Passing data to threads
- pthread_create
- pthread_exit
- pthread_self
- pthread_join
- pthread_attach
- pthread_detach
17